About shunt regs:
I found that this circuit could be optimal with the number of elements against chrs.
.
This current limiter prior shunt module performing significantly better then others.
I made a "test" with the spice. (Actually I was made real these shunts based on 431 and they all perform good vith the devices. But not this particular one.)
.
I found that this circuit could be optimal with the number of elements against chrs.
.
This current limiter prior shunt module performing significantly better then others.
I made a "test" with the spice. (Actually I was made real these shunts based on 431 and they all perform good vith the devices. But not this particular one.)
.
Attachments
The ESR value of output C is from the importance it should be not less than 0.1-0,15 ohms
otherwise some osculation may occur. I put the value of 47uF could be less say 22uF.
otherwise some osculation may occur. I put the value of 47uF could be less say 22uF.
As far as I'm concerned, the advantage of shunt regulators as compared to series regulators is that they have a lower transfer from load current variations to input current variations, that is, they keep the AC components of their load current more local. Have you also simulated the AC transfer from load current to input current?
I'm not able to contribute to the current topic of Agent.5's reclocking solution but am watching with interest from the perspective of doing a second Valve DAC build.
Apart from the possibilities around how you get DSD data into the Valve DAC, I'm also interested in exploring output options, such as the 'CD-Enhancer II' that Raj1 has brought to the party. Is there a downside to taking the output from the Valve DAC mainboard straight to XLR connectors and having the output stage solutions in separate boxes so they can be swapped easily?
Apart from the possibilities around how you get DSD data into the Valve DAC, I'm also interested in exploring output options, such as the 'CD-Enhancer II' that Raj1 has brought to the party. Is there a downside to taking the output from the Valve DAC mainboard straight to XLR connectors and having the output stage solutions in separate boxes so they can be swapped easily?
I have some thoughts,
I think 1Vrms from the Lundahl Filter PCB is too low to be optimal without a pre-amp in my system.
My current DAC has 2Vrms and it's OK. Would any of these balanced to single end buffers be an alternative to try if I need some more gain or just as an alternative to try against the transformer solution?
Tube buffer
GlassWare Unbalancer Two Kit
OP-amp buffer
Universal Buffer – Neurochrome
And as Ray wrote, having different buffer options in separate boxes would be a good idea if no downside, so one can easy experiment.
I also have some thoughts about a reclocking solution. Would Ian canadas FifoPi buffer be a working solution?
DocumentDownload/FifoPiUltimateManualV1.0.pdf at master * iancanada/DocumentDownload * GitHub
regards
Micke
I think 1Vrms from the Lundahl Filter PCB is too low to be optimal without a pre-amp in my system.
My current DAC has 2Vrms and it's OK. Would any of these balanced to single end buffers be an alternative to try if I need some more gain or just as an alternative to try against the transformer solution?
Tube buffer
GlassWare Unbalancer Two Kit
OP-amp buffer
Universal Buffer – Neurochrome
And as Ray wrote, having different buffer options in separate boxes would be a good idea if no downside, so one can easy experiment.
I also have some thoughts about a reclocking solution. Would Ian canadas FifoPi buffer be a working solution?
DocumentDownload/FifoPiUltimateManualV1.0.pdf at master * iancanada/DocumentDownload * GitHub
regards
Micke
Last edited:
As far as I'm concerned, the advantage of shunt regulators as compared to series regulators is that they have a lower transfer from load current variations to input current variations, that is, they keep the AC components of their load current more local. Have you also simulated the AC transfer from load current to input current?
I think i was not sim that I only did some ac transfer from input to the output to have some view into damping capability, FFT with same input to see if there are some resonance peaks or else, phase out (must say that for ALL shunts strat shifting phase very early), and output resistance that is uniformly very low for shunts. My say goal was to form circuit with smaller number of elements and affordable not so special parts.
.
I think that should be made one shunt and measure against the oscillations? Also maybe from higher importance, is to check thermal behavior BEFORE staring plot with PCB? Eventually employ for instance to supply that inverter input?
Some information's about shunt element and voltage reference, that could be interesting:
High-Current, Low-Voltage Shunt Regulator | Electronic Design
.
One topic at this forum was focused on pretty same circuit type but with some real life results.
Experimentations on Regulators
High-Current, Low-Voltage Shunt Regulator | Electronic Design
.
One topic at this forum was focused on pretty same circuit type but with some real life results.
Experimentations on Regulators
Is there a downside to taking the output from the Valve DAC mainboard straight to XLR connectors and having the output stage solutions in separate boxes so they can be swapped easily?
The downside is that the termination resistors no longer ground the output side of the DC blocking capacitors. That has two disadvantages:
1. In the original valve DAC, if due to some failure there should be -300 V at the input side of the DC blocking capacitors, touching the output could lead to a nasty shock. Fortunately the energy in the shock is limited, so it is not likely to kill anyone. In the raw DSD version, the mute JFETs will probably clamp the output voltage and prevent any shocks.
2. In the raw DSD valve DAC, when running without a filter connected, the voltage jump you get when switching between a DSD signal and an all zeros signal (for example due to a disconnected DSD input) might cause avalanche breakdown in the mute JFETs, which may or may not damage them.
If you want to go this way, connect a 47 kohm or 47.5 kohm resistor from each output pin to ground (so in total four resistors for a stereo DAC) and try never to use the DAC without one of the filters connected.
I have some thoughts,
I think 1Vrms from the Lundahl Filter PCB is too low to be optimal without a pre-amp in my system.
My current DAC has 2Vrms and it's OK.
I rather like the low level because it reduces the level difference with vinyl records, but anyway, it should be possible to use the Lundahl transformer as a 1:2 step-up. I'll get back to you about that.
Would any of these balanced to single end buffers be an alternative to try if I need some more gain or just as an alternative to try against the transformer solution?
Tube buffer
GlassWare Unbalancer Two Kit
This won't help, as it has < 0 dB voltage gain.
OP-amp buffer
Universal Buffer – Neurochrome
Same story, but the gain can be increased by changing resistors according to the site you linked to.
I also have some thoughts about a reclocking solution. Would Ian canadas FifoPi buffer be a working solution?
DocumentDownload/FifoPiUltimateManualV1.0.pdf at master * iancanada/DocumentDownload * GitHub
regards
Micke
It probably would, but as I don't see any timing specs, I can't be sure. If the output signals are TTL compatible and the data are stable from 5 ns before to 0 ns after the rising clock edge, then it should work when you install 45.1584 MHz and 49.152 MHz oscillators.
Hi Agent.5
"hand made" oscillators usually already have some sine to square module at the output? Please check with Your oscillator sch. Maybe it is not a good thing to repeat 2 X same circuit?
Yes, I will confirm, before finalizing the sine to square wave converter.
About shunt regs:
I found that this circuit could be optimal with the number of elements against chrs.
.
This current limiter prior shunt module performing significantly better then others.
I made a "test" with the spice. (Actually I was made real these shunts based on 431 and they all perform good vith the devices. But not this particular one.)
.
circuit and sim outputs looks good. Do I change R16 to adjust current out?
My equation is
R16 = Vbe / Io
R16 = 0.7V / Io
10Io = 0.7
Io = 0.07A = 70mA
circuit and sim outputs looks good. Do I change R16 to adjust current out?
Off-course adopt to Your need, this is an example only.
.
(I will try to find measured results of I consumption of ADUM1100 for I2S bus. As Higher Fo as higher consumption I remember. Maybe You can use some of this datas for your project).
But it is not the problem to measure with 1 adum isolator mounted on little PCB adapter, power with some 2 independent supplies, apply F digital signal and simple measure?
I have some thoughts,
I think 1Vrms from the Lundahl Filter PCB is too low to be optimal without a pre-amp in my system.
My current DAC has 2Vrms and it's OK. Would any of these balanced to single end buffers be an alternative to try if I need some more gain or just as an alternative to try against the transformer solution?
Tube buffer
GlassWare Unbalancer Two Kit
The original Broskie Unbalancer design offers gain;
Unbalancer
Yes, I will confirm, before finalizing the sine to square wave converter.
But You can see this from the oscillator circuit. At the end, most lightly, You already have the same buffer-sine-to-square module? Do You have Andrea's crystal clock boards?
Not yet. He is still testing and finalizing the result of the new group buy. I am pretty sure that it is sine wave. He is offering two different sine to square wave add-ons for Raspberry Pi.
Aha... Yes understood...
Probably I will order a piece. 🙂
.
(But I have already Andrea's and Ian boards for i think several years that i did not accomplished and set working yet 🙁
Maybe I shouldn't say that...)
Probably I will order a piece. 🙂
.
(But I have already Andrea's and Ian boards for i think several years that i did not accomplished and set working yet 🙁
Maybe I shouldn't say that...)
GB Update
The main DAC and reconstruction filter PCBs are more than 50% through the fabrication process so I think they'll probably ship by the end of this week.
The main DAC and reconstruction filter PCBs are more than 50% through the fabrication process so I think they'll probably ship by the end of this week.
I found a $40 solution that can be used to interface a BBB to the valve DAC. BBB can only accept 3.3VDC input and output. Therefore, we can not turbo-charge these isolators by running them at 5VDC.
The datasheet of these SI isolators are a bit confusing. They do not really say what voltage they are running at, just say 13ns max prop delay. If they are really 13ns at 3.3VDC, we can use them instead of the ADI 14ns at 3.3VDC chips and cut 2ns off a cycle. Anyway, SI8600 and SI8605 are slow. But you get SI8621, 8655, and 2 8663.
SI8621 has 1 forward and 1 reverse channel. SI8655 has 5 forward channels and 0 reverse channel. SI8663 has 3 forward channels and 3 reverse channels.
Si86xx Isolator Evaluation Kit - Silicon Labs
The datasheet of these SI isolators are a bit confusing. They do not really say what voltage they are running at, just say 13ns max prop delay. If they are really 13ns at 3.3VDC, we can use them instead of the ADI 14ns at 3.3VDC chips and cut 2ns off a cycle. Anyway, SI8600 and SI8605 are slow. But you get SI8621, 8655, and 2 8663.
SI8621 has 1 forward and 1 reverse channel. SI8655 has 5 forward channels and 0 reverse channel. SI8663 has 3 forward channels and 3 reverse channels.
Si86xx Isolator Evaluation Kit - Silicon Labs
Last edited:
Table 4.3 of https://www.silabs.com/documents/public/data-sheets/Si866x.pdf applies at 3.3 V +/- 0.33 V supply voltage and states that the maximum delay is 13 ns.
- Home
- Source & Line
- Digital Line Level
- Valve DAC from Linear Audio volume 13